Any problems using a switch on the lamp side of a Ballast?

I have a 18 W low pressure sodium lamp that is powered by a magnetic ballast. Looks like a simple inductor with a power correction capacitor (photo from old prototype box):

The lamp is then on about 4m of cord to a pendant floor lamp (heh, who says LPS isn't suitable for indoor lighting), that happens to have a inline switch.

           switch                 
  |-------/+-------------_______---A
  |                 ----|Ballast|
Lamp           Cap  |   |_______|
  |                 =
  |-------/+--------|------------- N

Now with this switched off the lamp is effectively taken out of the circuit, but the Ballast and Capacitor are still powered. Is this a problem?

I have trialled it. The Ballast stays cold. However if you use a cheap Power-Watt meter it says it's using about 80Watt! power factor 100%.
This is in contrast to normal running conditions where it reads 25 Watt, 97% power factor. I tried looking at the power meter in the switch box, but on/off/unplugged seemed to have no visually measureable effect.

 

At the very least, the capacitor continues to have mains across it and constitutes a reliability (fire) risk.

I 'm not that familiar with LPS, but think there may be a worse problem. Are not the choke and capacitor arranged so as to increase the available voltage for starting, and/or to compensate for rising arc voltage as the tube ages? In this case, with the tube out of circuit there could be be a large voltage developed at the output of the ballast. The choke and the capacitor would be highly stressed, anyone handling the switch might get a shock if the insulation failed due to the excessive voltage.

Maybe someone else will be able to rule this out, but it sounds very risky to me. Better not do it unless you can be sure.

 

Fair point!
I know one end of the capacitor is grounded, and the other lead goes to the point labeled C for capacitor on the ballast in the picture. Not so sure where this in relation to the windings and such.

I thought the capacitor is purely a power correction thing.

I gather LPS is electrically the same as (or very similar to) fluorescent tubes. The small ones (usually 18W like this) don't need a starter, just like the small fluoro tubes that are in desk lights and such.

I do have a desk light that has a similar set up with a 11W PL tube with a switch on the cord to a simple wall wart iron ballast. It's iron ballast remains cold, with minimal electromagnetic field (by magnet) when the switch to the tube is off. But perhaps it has no capacitor...? I once measured one of these; open circuit voltage was about 230 Volt. So switch failure would be no more dangerous than by standard full mains table lamp switch.

I think I'll go open it up again after a bit powered on but switched out. And check the capacitor isn't unusually warm.

But what happened to these things when the lamp fails or is missing??

 

Hmm.
There's no buzz when holding a magnet up to the ballast, so doesn't seem to be drawing any noticeable current. On opening it now, capacitor feels cold.

And certainly not using the 80 W the cheapo power meter claims. Must be a quirk of the inductive coil.

On safety side of things: The box is earthed at two points and has a fuse on the active line, so i guess if the capacitor does go kaput shouldn't do anything devastating.

 

Use the wall switch not the in-line.

 

Yeah, agreed. I think I'll change the lead for one without a switch too.